Determining redox properties of clay-rich sedimentary deposits in the context of performance assessment of radioactive waste repositories: Conceptual and practical aspects

Redox reactions play a key factor controlling the mobility of redox sensitive radionuclides in clay-rich sediments which might serve as host formations for radioactive waste repositories. Assessing the redox speciation of radionuclides requires information about the redox conditions in the formation and the electron transfer kinetics between the redox active constituents in the sediment and the radionuclides. Clay-rich sediments, like most natural systems, are usually not in redox-equilibrium when they received sufficient organic material during deposition to enforce anaerobic conditions during early diagenesis. Consequently, any derived redox potential (Eh) to characterize the conditions in the formation should include an evaluation about which half reactions were probed by the used technique or indicate for which half reaction(s) the value was calculated. Conventional Eh measurements, performed with an inert electrode in the pore water retrieved from clay-rich sediments, are unlikely to provide meaningful values. This is due to the intrinsic limitations of this technique to probe the redox potential of environmentally relevant dissolved constituents and because many redox reactions in clay-rich sediments involve solids. Electrochemical techniques which probe the redox properties of solids in aqueous suspensions or in salt matrix are promising alternatives but their applicability in clay-rich sediments still has to be tested. When these techniques live up to the expectations, their application in combination with geochemical characterization and thermodynamic calculations might be the most reliable approach to constrain the Eh of redox active constituents in clay-rich sediments. A large variety of redox active species of the elements H, O, N, C, Fe, Mn and S might be present in clay-rich sediments. A straightforward approach to obtain Eh values for various redox couples in the sediment is to use thermodynamic calculations in combination with a rigorous geochemical characterization of the sediment. However, not for all redox active species, which are of possible importance for controlling the redox states of radionuclides, reliable thermodynamic data are available. This is in particular the case for structurally bound Fe in clay minerals, organic matter, and secondary minerals formed during early diagenesis or after burial. The role of many of these redox couples for redox transformations of radionuclides has been experimentally investigated under environmentally relevant conditions. Examples for different reactions are compiled in this report. In general, the reduction of oxidized species of U, Se, and Tc by reduced species of Fe, S, and C in the absence or presence of microbial mediation have been frequently studied. Studies on transuranium elements, other redox sensitive radionuclides, and the oxidation of radionuclides are more rare. Information about the kinetics of redox reactions between radionuclides and the oxidized or reduced species of redox sensitive major elements in the formation should be used to interpret the Eh value(s) obtained for one or more redox couples in the formation. An Eh value obtained for a redox couple in the sediment is only of relevance for predicting the redox state of a radionuclide when the time scale of the redox reaction is shorter than the half-life of the radionuclide and the residence time of the radionuclide in the formation. Often, redox reactions between radionuclides and redox active constituents of the clay-rich sediment only proceed rapidly in one direction while the reaction in opposite direction is kinetically hindered. In this case, the Eh derived for a redox couple in the sediment should only be used as an upper or lower limit to constrain the redox speciation of the radionuclide of interest. As a general conclusion, it can be stated that while clay-rich sediments in many cases are considered to be reducing, the exact redox potential, the in situ activity of the various redox couples, and their effect on redox sensitive radionuclides are less evident. This frequently results in pragmatic and stylised approaches when transferring phenomenological knowledge to safety assessment calculations